APFIM Characterization of Grain Boundary Segregation in Titanium Carbide-Doped Molybdenum

Transcription

1 APFIM Characterization of Grain Boundary Segregation in Titanium Carbide-Doped Molybdenum M. Miller, H. Kurishita To cite this version: M. Miller, H. Kurishita. APFIM Characterization of Grain Boundary Segregation in Titanium Carbide-Doped Molybdenum. Journal de Physique IV Colloque, 1996, 06 (C5), pp.c5-265-c < /jp4: >. <jpa > HAL Id: jpa Submitted on 1 Jan 1996 HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

2 JOURNAL DE PHYSIQUE IV Colloque C5, supplkment au Journal de Physique 111, Volume 6, septembre 1996 APFIM Characterization of Grain Boundary Segregation in Titanium Carbide-Doped Molybdenum M.K. Miller and H. Kurishita* Microscopy and Microanalytical Sciences Group, Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN , U.S.A. * Institute for Materials Research, Tohoku University, Ibaraki, , Japan Abstract. The grain boundary segregation behavior of titanium and carbon have been characterized in two titanium carbide-doped molybdenum alloys. The matrix of these alloys was found to be significantly depleted in titanium, carbon, oxygen, and nitrogen. Both titanium oxycarbide and molydenum carbide precipitates were observed. The Gibbsian Interfacial Excess, Ti, determined from atom probe analyses revealed significant enrichments of carbon and nitrogen at the grain boundaries in both alloys. In atom probe analyses where the carbon level was high at the grain boundary, the oxygen level was below the detection level and donversely, in analyses where the carbon level was low, some oxygen was detected and suggests that carbon has a beneficial effect of displacing the oxygen from the grain boundary. 1. INTRODUCTION Recrystallization embrittlement and irradiation embrittlement are major problems for materials to be used in fusion reactors. The severe environment in plasma-facing components in a fusion reactor consists of high heat loads and irradiation by high energy particles. Molybdenum has several attractive properties for high temperature structural applications including high melting point, good thermal conductivity, high resistance to sputtering and erosion, and high strength. However, molybdenum alloys also exhibit brittleness at low temperature [l-31. This brittleness is associated with intergranular fracture which may be an intrinsic property of molybdenum but can be promoted by the segregation of interstitial impurities such as oxygen and nitrogen to grain boundaries at low temperatures due to their low matrix solubilities. Commercial molybdenum alloys such as TZM (MO- 0.5% Ti- 0.1% Zr) exhibit good low temperature toughness but after recrystallization, their ductile-to-brittle transition temperatures are above room temperature and they suffer from significant irradiation embrittlement. The addition of titanium carbide to molybdenum has been found to strengthen intrinsically weak grain boundaries and thereby lower the ductile-to-brittle transition temperature. In a previous study, Kurishita et al. demonstrated that a Tic-doped MO alloy had a lower ductile-to-brittle transition temperature and significantly higher strength than a commercial TZM material, in the as-rolled, stress relieved, recrystallized, and neutron irradiated conditions 121. However, there is little direct evidence of the roles of titanium and carbon and their interaction with oxygen in this Tic-doped material. The energy-compensated atom probe field ion microscope has been previously demonstrated to be an ideal tool for this type of high resolution characterization due to its light element sensistivity and high mass resolution. The suitability of the atom probe for grain boundary characterizations has been reported previously [4]. Therefore, an atom probe field ion microscopy characterization of the solute distribution in Tic-doped molybdenum has been performed. The presence of oxygen at grain boundaries in molybdenum was experimentally demonstrated by Waugh et al. with the use of imaging atom probe elemental maps [5, 61. An average oxygen concentration of approximately 2% was measured in a 1 nm thick region of the grain boundary. This initial result was later confirmed by conventional atom probe analysis of a number of grain boundaries [7]. Article published online by EDP Sciences and available at

3 CS-266 JOURNAL DE PHYSIQUE IV 2. EXPERIMENTAL The grain boundary segregation behavior of titanium and carbon have been characterized in two molybdenum alloys. The alloys were prepared by powder metallurgy methods with mechanical alloying of the molybdenum (99.9% purity) and Tic powders (98% purity) and hot isostatic pressing treatments. The initial size of the molybdenum powder was 5 pm and that of Tic powder was 0.57 pm, but they were reduced to -30 nm after the mechanical alloying in an argon atmosphere and hot isostatic pressing treatments. Mechanical alloying was performed at ambient temperature and the hot isostatic pressing treatment was performed at K at 200 MPa in argon. The density of the material after hot isostatic pressing was 99.8% of the theoretical density. The densified body was then forged at 1500 K and finally hot (-1500 K) and warm (-700 K) rolled into 1 mm thick sheets. Specimens for transmission electron microscopy and atom probe field ion microscopy were cut from these sheets. The compositions of these two alloys are given in Table 1. In addition to the elements listed in Table 1, the MTC-02 alloy contained 56 pprn of Si, 35 pprn of Fe, 21 pprn of Ni, <20 pprn of CO, Ca, Cr, Mg, and < 5 pprn of Al, Cu, Mn, Pb and Sn; MTClO contained 65 pprn of CO, 42 pprn of Si, 29 pprn of Fe, < 20 pprn of Cr and Ni, and < 5 pprn of Ca, Al, Cu, Mg, Mn, Pb and Sn. Table 1. Compositions of the alloys used in this study. The balance is molybdenum. Alloy MTC-02 MTC-10 wppm at. % wppm at. % Ti C N W The characterization was performed in the ORNL energy-compensated atom probe field ion microscope [g] with a specimen temperature of 50 K, a pulse fraction of 20% and an analysis pressure of 5 X 10-lo mbar. The base pressure of the instrument before analysis was -2 X 10-l' mbar. Helium was used as the image gas. This high mass resolution instrument is able to minimize some of the problems of peak overlap in the mass spectrum due to Mo4+/Ti2+ peaks at 23 to 25 u and the 160+ l4'ti3' peaks at 16 U. Transmission electron microscopy characterization was performed with a JEM-200CX and a JEM- 4000FX operated at 200 and 400 kev, respectively. 3. RESULTS AND DISCUSSION Transmission electron microscopy revealed the presence of some ultrafine precipitates at the grain boundaries [3], as shown in Fig. l. The size of these precipitates was less than 60 nm [3]. Energy-dispersive X-ray (EDS) analysis of these particles revealed that they were Ti-oxycarbides. The grain size of the material varied from 10 to 400 nm and the grains were elongated along the rolling direction. This fine grain size is significantly smaller than that of traditional alloys and may have a major effect on the mechanical properties. This small grain size significantly increases the quantity of grain boundary area in the material that is available for solute segregation and may therefore reduce the solute coverage at any point on the grain boundary. EDS analysis of the grain boundary segments between these particles revealed enrichments of both titanium and oxygen. However, these EDS results should be treated with some caution since the Ti, peaks overlap the oxygen peaks and the MO, peaks overlap the carbon peaks. Despite extensive field evaporation of many specimens, only one precipitate was observed by field ion microscopy. A field ion micrograph of this coarse brightly-imaging precipitate in the MTC02 alloy is shown in Fig. 2. The precipitate was larger than the field of view of the field ion microscope and was therefore >l00 nm. The composition of this coarse precipitate was determined by atom probe field ion microscopy and found to be 66.0 f 1.0 at. % MO, 32.0 f l.o% C, 1.2 f 0.05 % N. From peak deconvolution, the maximum titanium and oxygen levels in this precipitate were determined to be 0.6% and 0.08%, respectively. The composition of the precipitate was consistent with Mo2C. The carbon-to-nitrogen ratio was 27: 1. This ratio is significantly higher than the ratio (3.5:l) in the alloy. Since the carbon was added in the form of Tic powder, the presence of these Mo2C precipitates indicated that the Tic powder has dissolved and suggests that the titanium content of the matrix should be higher than the carbon content.

4 Atom probe composition determination in this material is complicated by isobar overlaps and the low solute concentrations of the minor elements. Under the ex erimental conditions used, molybdenum field evaporates as Mo3', MO'' and a small quantity (<2%) of MO bl. Despite the low background pressure during analysis, He+, MOH~~', MOH~~' and Ne' species were observed in matrix analyses. Titanium generally field evaporates as ~ i and ~ ' possibly Ti2+. Therefore, all five Ti2' peaks are exactly coincident with the corresponding Mob peaks at u and only the 95Mohand 97Mob are free of ambiguity. Examination of the mass spectra of the matrix in the 23 to 25 u range revealed that the seven observed peaks fit the natural abundances of molybdenum to within the statistical scatter. Therefore, the titanium concentration was taken as only the Ti3' peaks. There is also an overlap between the 160+ and the 48~i3' peaks. Since no 0' ions were detected, the oxygen content is quoted as the summation of all the ions detected at 16 U. The ox en value therefore represents the maximum possible concentration as the 16 u peak will also contain some '?i3+ ions. No Tic, MoC, MoN, MOO, or MOO, species were observed in the matrix. No significant peak overlaps are present in the atom probe spectra for carbon and nitrogen. The matrix compositions for the two alloys determined from atom probe analyses are summarized in Table 2. These results indicate that the matrices of both alloys are significantly depleted in titanium, carbon, oxygen and nitrogen compared to the bulk composition. The results also indicate that these elements are either associated with precipitates or segregated to other microstructural features such as grain boundaries. These results are in agreement with the low solubilities of carbon and titanium in molybdenum [9]. Table 2. Matrix compositions determined from atom probe analyses. The balances of these analyses are molybdenum. The results are given in atomic percent. Element Ti 0 C N MTC f max rf: rf: MTC f max f not detected Field ion micrographs revealed a relatively low coverage of bright spot decoration along the grain boundaries in both MTCO2 and MTClO alloys, as shown in Figs. 3 and 4, respectively. Although atom probe single atom catching experiments were performed to determine the identity of these bright spots, the majority of the bright spots were found to be molybdenum atoms. This result suggest that molybdenum atoms were displaced from their lattice sites. No clear evidence of any precipitates along the grain boundaries were observed in any of the grain boundary segments encountered during field evaporation. Several atom probe analyses were performed on the grain boundaries in both materials to determine the segregation behaviour of the solutes. The level of segregation was evaluated from the atom probe data with the Gibbsian interfacial excess method [10]. The Gibbsian interfacial excess, Ti is given by [10,1 l] ri = N,(excess)lA = (N, - Ni(a) - Ni(b))lA, where N i(e,,g is the excess number of solute atoms associated with the interface, A is the interfacial area over which the lnte acial excess is determined, N, is total number of solute atoms in the volume analyzed, N,(,) and Ni(b) are the number of solute atoms in the two adjoining regions a and b either side of the dividing surface. However, since the matrix solubilities of the solutes are so low in this material, the values of N,(,) and NI(,, were always zero and therefore, the Gibbsian interfacial excess reduces to ri = N,/A The Gibbs~an interfacd excess determined from the atom probe analyses are summarized in Table 3. In these analyses, the oxygen was taken as the summation of all the ions in the 16 u peak. Therefore, the quoted values represent an upper limit for the oxygen segregation. The titanium segregation behaviour could not be reliably estimated from most of these individual analyses due to the small number of solute atoms involved and the overlap of the Ti2' isobars with the Mob peaks. In order to relate these values to monolayer coverages, a definition of a monolayer must be used. One method is to calculate the number of atoms expected on the closest packed plane of the crystal structure. For body centered cubic molybdenum, this is the (110) plane and therefore a monolayer coverage is 1.40 X 1019 atoms m-2 with a width W = d(110) = 0.22 nm. The monolayer coverages, 0, were estimated based on a saturation value of = 1.40 X 1015 atoms cm-2. An alternative method is to use a more practical definition in which 1 monolayer contains a-2 segregant atoms per unit area where a3 is the atomic volume of the segregant [12]. If this method is adopted, the coverages are decreased by approximately 7% from the values quoted in Table 3.

5 C5-268 JOURNAL DE PHYSIQUE IV Figure 1: Transmission electron micrograph showing fine pre- Figure 2: Field ion micrograph of a coarse Mo2C precipitate in cipitates (c) along a grain boundary in the as-rolled MTC-02 alloy. the as-rolled ~ ~ c -alloy, 0 2 1( f Figure 3. F~eld Ion rn~cro~raphh hhowlnp dccor.~t~on along pram mdmes 111 IIIC as-rolled MTC-0'2 alloy. Figure 1. Field Ion m~crogmphh show~ng decorallon along pram boundaries in clle as-rolled IMTC-I0 alloy

6 The results revealed enrichments of carbon and nitrogen at the grain boundaries in both alloys. No significant difference was observed between the two alloys. The carbon-to-nitrogen ratio was approximately 5:l in both alloys and therefore similar to the ratio for the bulk material. In most analyses, the oxygen level was found to be below the detection level of the atom probe analysis. However, oxygen enrichment was observed in some regions. It is interesting to note that in all the atom probe analyses where the carbon level was high (i.e., greater than -6%), the oxygen level was below the detection level and conversely, in analyses where the carbon level was low (i.e., <4%), some oxygen was detected. This preliminary result would suggest that carbon has a beneficial effect of displacing the oxygen from the grain boundary, thereby reducing the problem of oxygen embrittlement. Table 3. The Gibbsian Interfacial Excess, ri, determined from atom probe analyses. MTC Average MTC Average r, loi4 atoms cm-2 0 (max.) C N n.d. 1.O 0.1 n.d. 1.6 n.d. n.d. 1.1 n.d O C N n.d O n.d % 0 (max.) C N n.d n.d n.d. n.d. 8.0 n.d C N n.d S n.d. 0.8 The atom probe results appear to be different from that measured in the transmission electron microscope. However, it should be noted that the value determined in the transmission electron microscope was obtained from a grain boundary segment in between and in close proximity of two titanium oxycarbide precipitates, whereas a11 the atom probe analyses were performed at grain boundary segments removed from precipitates. Therefore, it is possible that the grain boundary precipitates can act as a source of oxygen and a sink for carbon and thereby alter the local composition of the grain boundary. Further atom probe and transmission electron microscopy research is required to clarify the role of these grain boundary precipitates. 4. CONCLUSIONS The matrix of these Tic-doped molybdenum alloys was found to be significantly depleted in titanium, carbon, oxygen, and nitrogen. Both titanium oxycarbide and molydenum carbide precipitates were observed. The Gibbsian Interfacial Excess, Ti, determined from atom probe analyses revealed significant enrichments of carbon and nitrogen at the grain boundaries in both alloys. In atom probe analyses where the carbon level was high at the grain boundary, the oxygen level was below the detection level and conversely, in analyses where the carbon level was low, some oxygen was detected and suggests that carbon has a beneficial effect of displacing the oxygen from the grain boundary. Acknowledgments The authors would like to thank K. F. Russell for her assistance and Drs. J. Bentley and E. A. Kenik for helpful discussions. This research was sponsored by the Division of Materials Sciences, U. S. Department of Energy, under contract DE-ACOS-960R22464 with Lockheed Martin Energy Research Corp. and by Grant-in- Aid for Scientific Research (B) (# ), the Ministry of Education, Science and Culture, Japan. This research was conducted utilizing the Shared Research Equipment (SHaRE) User Program facilities at Oak Ridge National Laboratory.

7 C5-270 JOURNAL DE PHYSIQUE IV References [l] Kurishita H., Asayama M., Tokunaga 0 and Yoshinaga H., Mater. Trans. JIM, 30 (1989) [2] Kurishita H., Kitsunai Y., Hiraoka Y., Shibayama T. and Kayano H., Mater. Trans. JIM, 37 (1996) [3] Kurishita H. and Yoshina a H., Materials Forum, 13 (1989) Miller M.K. and Smith &.W., AppL Surf Sci., 87/88 (1995) [5] Waugh A.R. and Southon M.J., Surf Sci., 58 (1977) [6] Waugh A.R., J. Phys. E: Sci. Instrum., 11 (1978) [7] Waugh A.R. and Southon M.J., Surf Sci., 89 (1979) [S] Miller M.K. J. de Phys., 47-C2 (1986) [9] e.g., Massalski T.B., ed., "Binary Alloy Phase Diagrams," (American Society for Metals, Metals Park OH, 1986) Gibbs J.W., "The Collected Works of J. Willard Gibbs," vol. I, (Yale University Press, New Haven, CT, 1948). [I l] Lupis C. H. P., "Chemical Thermodynamics of Materials," (Elsevier Science, New York 1983) pp [l21 Hondros E.D. and Seah M.P., Inter. Mater. Rev., 22 (1977)